There are an estimated 250,000 sudden cardiac deaths (SCD) annually in the United States constituting approximately 50% of all cardiac deaths. Although clinical trials have demonstrated convincing survival benefits conferred by implantable cardioverter defibrillator (ICD) therapy in selected patients with left ventricular ejection fractions (LVEF) less than 35% and congestive heart failure, the overwhelming majority of patients who suffer a cardiac arrest will have an LVEF> 0.35. In this competing continuation of the PRE- DETERMINE: Biologic Markers and MRI SCD Cohort Study, we propose to determine whether biologic markers and ECGs can be utilized to advance SCD risk prediction in patients with CHD and LVEF>30-35% where methods for SCD risk prediction are lacking. In the first grant cycle, 5764 patients with CHD and LVEF>30-35% were enrolled in multicenter prospective cohort study, PREDETERMINE, and blood samples and ECGs were collected, stored, and a panel of biomarkers, fatty acids, and a multitude of ECG parameters have been/will be measured. In addition, a wealth of information on demographics, clinical and lifestyle characteristics, and cardiac test results has been collected and stored. In a subset of patients, contrast- enhanced cardiac MRIs (CE-MRI) were collected and analyzed for infarct size and other morphologic features. The cohort has been closely followed centrally by mail and phone for 3 years with low rates of lost to follow-up, and adjudication of cause specific mortality and arrhythmic events is ongoing. The primary endpoint is a combined endpoint of sudden and/or arrhythmic death (SAD) and out-of-hospital VF arrest (VF). In the Competing Continuation, we will leverage this rich and unique resource created during the first grant cycle to address the following aims regarding SCD risk prediction in this understudied population. 1).To evaluate whether biomarkers and ECG characteristics can be utilized to identify the presence of high risk arrhythmogenic myocardial scar on CE-MRI. 2).To determine whether these and other biologic and ECG markers associated with SCD in the general population are associated with SAD/VF in this population. 3).To develop clinically useful predictive models based on combinations of biomarkers, ECG characteristics and conventional risk factors that predict risk of SAD/VF as opposed to other causes of mortality in patients with CHD who do not have severe systolic dysfunction. 4).To test whether genetic risk scores can add to SAD/VF risk reclassification in CHD patients who do not have severe systolic dysfunction. If biomarkers, ECG, or genetic markers are identified that can predict the occurrence of SAD/VF to a greater extent than other causes of mortality in this population, then these markers may serve as relatively inexpensive methods to identify patients with CHD and LVEF>30-35% who might benefit from the ICD. The findings may also enhance our mechanistic understanding of SAD in the setting of CHD, which could lead to novel preventive approaches in the general population, where CHD underlies the majority of SCD.